Genetic Programming-Based Feature Selection for Emotion Classification Using EEG Signal.

Area of Science:

• Neuroscience
• Artificial Intelligence
• Mental Health Research

Background:

• The COVID-19 pandemic and associated lockdowns have significantly impacted global mental health.
• Emotion recognition is crucial for early diagnosis and effective treatment of mental health issues.
• Artificial intelligence, particularly genetic programming (GP), shows promise in addressing complex research challenges.

Purpose of the Study:

• To propose a novel genetic program-based feature selection (FSGP) technique for emotion recognition.
• To identify and select relevant features from electroencephalogram (EEG) data for improved classification accuracy.
• To enhance the efficacy of mental health diagnostics through advanced AI methods.

Main Methods:

• Utilized a fourteen-channel electroencephalogram (EEG) device to acquire brain signal data.
• Applied a genetic programming (GP) based feature selection (FSGP) technique to process 70 initial features.
• Reduced the feature set by separating irrelevant and redundant data, selecting 32 key features.

Main Results:

• The proposed FSGP model successfully selected 32 relevant features from the initial 70.
• Achieved a classification accuracy of 85% for emotion recognition using the selected features.
• Demonstrated superior performance compared to existing methods in prior research.

Conclusions:

• The FSGP technique is effective in optimizing feature selection for emotion recognition from EEG data.
• This approach holds significant potential for improving the accuracy and efficiency of mental health diagnostics.
• Further research in AI-driven emotion recognition can lead to better mental healthcare solutions.